You are on page 1of 9

Available online at www.sciencedirect.

com

Procedia Technology 5 (2012) 190 198

CENTERIS 2012 - Conference on ENTERprise Information Systems / HCIST 2012 - International


Conference on Health and Social Care Information Systems and Technologies

A solution for real time monitoring and auditing of


organizational transactions
Rui Pedro Marquesa,*, Henrique Santosa, Carlos Santosb
a
Algoritmi, Universidade do Minho, Campus de Azurm 4800 Guimares, Portugal
b
Govcopp, Universidade de Aveiro, Campus de Santiago 3810 Aveiro, Portugal

Abstract

The controlling and auditing of organizational transactions in real time allows to determine the degree of reliability with
which they are carried out, mitigating the organizational risk. This paper presents a solution proposal under a new vision
for organizational auditing and monitoring in real time since it is focused on the implementation of continuous assurance
services in organizational transactions in compliance with the formalisms of a business ontological model. Furthermore,
this paper contributes for a new paradigm of the transactional auditing, which is intended to be at a very low and detailed
level of organizational transactions.

2012 Published
2012 by Elsevier Ltd.
Published SelectionLtd.
by Elsevier and/or peer review
Selection under
and/or responsibility
peer-review of CENTERIS/SCIKA
under responsibility of -
Association for Promotion and Dissemination of Scientific Knowledge Open access under CC BY-NC-ND license.
CENTERIS/HCIST.

Keywords: organizational transactions; management information system; continuous assurance; monitoring; auditing; real time; risk
profiles;

* Corresponding author. Tel.: + 351 253 510180; fax: + 351 253 510189.
E-mail address: ruimarques@ua.pt .

2212-0173 2012 Published by Elsevier Ltd. Selection and/or peer review under responsibility of CENTERIS/SCIKA - Association for Promotion
and Dissemination of Scientific Knowledge Open access under CC BY-NC-ND license.
doi:10.1016/j.protcy.2012.09.021
Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198 191

1. Introduction

Currently organizations are facing several challenges, for example their organizational transactions have
grown in volume and complexity and they are living in highly regulated business environments. Thus,
controlling and monitoring mechanisms are needed in order to evaluate and validate all transactions, in a
comprehensive manner, to meet the controls and regulations. However, the traditional audit process occurs
mostly after the completion of transactions, since it is not feasible to audit them in time. Thereby it makes it
possible to inhibit the risk associated to their execution. Therefore, for many organizations there is a
significant risk of errors and fraud and these are not detected in time, resulting in a negative impact on
organizations. See, for example, the current global financial crisis and successive well-known scandals in
some organizations, such as Lehman Brothers, A-Tec, Madoff, Kaupthing Bank, WorldCom, Enron, Parmalat
and Tyco cases and many others [1-4].
Thus, any organization must be sufficiently prepared to survive, regardless of exposure and of the large
number of risks it is subject to, by implementing a suitable system of Continuous Assurance in accordance
with applicable legislative and regulatory framework. Continuous Assurance has been assuming an important
role within the organizational context because it is the application of emerging information technologies to the
standard techniques of auditing. "Continuous" does not mean real time, but it means to be effective,
considering and being consistent with the pulse and rhythm of each organizational transaction and process [5,
6].
These aspects have propelled to create a new awareness of corporate governance and of the growing
importance of monitoring and controlling the various organizational transactions (that is, any activity
performed within a business process). Along with this evidence, a study by PricewaterhouseCoopers [7]
examined various organizations and concluded that about 89% of participating organizations intend to adopt
more solutions of continuous auditing and monitoring by 2012.

1.1. Motivation

Given the foregoing, it is necessary to find solutions which allow organizations to evaluate, monitor and
validate their transactions continuously and independently, preferably in a non-intrusive way. The
optimization of the operational performance will be also possible if this auditing is done in real time (in the
shortest time possible after its execution), reducing in this way the associated risks.
Alongside this, there is another aspect to consider in relation to organizational transactions: risk profiles. In
this context, risk profiles refer to the classification of different types of behavior that may occur in the
execution of one transaction. In this work, two terms are considered to characterize risk profiles: negative
profiles, which refer to all unwanted behaviors during the execution of transactions, for example incomplete or
poorly executed operations; lack of crucial procedures; non-conformities; delays; incongruities and
malfeasance and positive profiles, which refer to all valid and appropriate events [8, 9].
Thus, this paper focuses on answering to the implementation of real time assurance services, having as
support the organizational transactions according to an ontological model of organizational transactions. An
ontological model is important because it helps to understand the essence of the organizational transactions
and processes and their relationships and characteristics. In parallel, a simpler business view, detached from
any ontological representation, results in the inability to generate organizational knowledge [10]. Therefore,
this work intends its prototype to be a system with a broader and detailed vision of organizational processes
and transactions, thus respecting the formalities of an ontological model capable of representing the
organizational reality. The work presented in this paper is supported by Enterprise Ontology, the model
proposed by Dietz [11].This model is adapted to represent the essential structure of the organizational
192 Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198

transactions with no significant complexities but simultaneously with coherence (i.e. parts constituting an
integral whole), consistency (there is no contradiction or irregularities), comprehension (all the important
issues are handled) and concisely (i.e. model does not contain superfluous matters). Furthermore, it has been
applied successfully in some practical projects in recent years [12].
This paper is structured in six sections, including this introduction to the topic and the motivation. In the
next section follows a brief literature review. Then, the solution proposal is presented, one that shows
evidence of being effective in achieving the stated objectives. Section 4 presents the main results which are
intended to be achieved and overviews a methodology for implementation and evaluation of the proposed
solution. Finally, the last section presents the authors conclusions.

2. Literature review

This section aims to present some paradigms and concepts associated with the monitoring and auditing at
the level of execution of the organizational transactions and processes. Some researches and applications
related to the topic are also presented. Thus, the concepts which are here evinced have features and
specificities similar to those presented for the prototype conceptualized in this paper.
The first two concepts intended to refer are Business Activity Monitoring (BAM) and Business Process
Management (BPM). BAM allows that events generated by various applications and systems of an
organization, or by services of inter-organizational cooperation, can be processed in real time in order to
identify critical situations in the performance indicators. It aims to obtain a better insight into the business
activities, and thus improve their effectiveness. BAM identifies and analyzes in real time the cause-effect
relationships between events, enabling the system and/ or staff to take effective and proactive measures in
response to specific identified scenarios. It allows, for example the early detection of abnormal events in
business processes as a whole, or some of their constituent parts [13, 14]. In turn, BPM is defined as
supporting business processes using methods, techniques and software to design, enact, control and analyze
operational processes involving humans, organizations, applications, documents and other sources of
information [15]. The BPM systems are complex assemblies of software components and tools that together
provide features which allow us to develop, deploy and implement solutions based on business processes.
Moreover, they enable the visualization, monitoring and management of events within the business process,
and allow the highest-level visualization of the state of the execution of business processes, reducing the
causes of the occurrences of exceptions [14].
Another interesting concept to present here is Complex Event Processing, (CEP) which includes methods,
techniques and tools to process events in real time. CEP analyzes a series of data in real time and identifies
patterns and generates events that can be processed and treated. This processing is done in memory and its
logic is defined by a series of queries on all received data [16, 17]. In short, CEP is capable of processing high
amount of data from different sources; operates in bitstream; has low latency; has limited processing window
and can handle different types of operations on data, such as filtering, correlation, aggregation and association
patterns.

2.1. Related work

Researchers from the Brandeis University, Brown University, and MIT carried out the project Aurora [18].
This project was designed to handle and manage a very large amount of data streams and allows its users to
create their own queries from a set of available operators. These operators are connected to other operators or
may simply provide results. These operators may derive from the output data from other operators or external
data sources. Aurora is capable of optimizing the query considering the QoS indexes provided by the operators
Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198 193

and other indexes and system inputs specified by the users. This was a precursor of other identical monitoring
systems of data flow, e.g. Medusa [19] and Borealis [20]. Like Aurora, the project STREAM [21] is a Data
Stream Management System. The STREAM supports a large number of declarative continuous queries over
continuous data streams and/ or over traditional data repositories. The monitoring is done by controlling the
results of queries made.
The work EasyCredit [14, 22] is an example of successful implementation in the banking sector. It is a
system like BAM, using the concept of CEP and the pipeline used in the management and monitoring of credit
transactions in real time.
Some works on this topic, in which event monitoring is done using log records, were also found. Within
this group of works surveyed, one of them uses mining tools and CEP to analyze records of a database log in
real time, and presents the sequence and the model of the transactions analyzed [16]. Furthermore, using
mining techniques on log records, the possibility of recognition of events was demonstrated. It is able to link
events that are not associated a priori with any workflow or process model to a new model, in other words it
contributes for the discovery of new process and transactional models [23].
Another work of reference [24] describes an approach for automating the discovery of patterns of activity
in organizational process models through an ontology. This discovery of patterns is done through a mapping
between the elements collected in the processing of the process and the elements of the ontology. This is done
primarily to verify if the process contains the necessary elements to meet the definition of each process
pattern.

3. Solution proposal

This chapter aims to present a proposal for a solution which shows evidence of being effective in achieving
the stated objectives. In this proposal the requirements that must be implemented are presented along with a
possible conceptual architecture to meet these requirements. However, this section does not aim to describe
technologically the solution, but to clarify, from a conceptual point of view, the purpose of each one and their
relationships, following the alignment of the requirements which will be described.
A solution that addresses the problems and motivations presented in this paper should meet some
requirements. The first one is the necessity to conceptualize a layer of non-intrusive internal control
mechanisms in order to be incorporated in the operational information system (e.g. in the ERP). This system
supports the execution of the organizational transactions to be monitored and audited. These internal control
mechanisms, when embedded into ERP system, must be aligned with the ontological model of Dietz. In other
words, the design of these mechanisms will have to take into account the different types of events, stages and
relationships that constitute the essence of each transaction. Furthermore, a component that manages and
stores the results which derive from the internal control mechanisms is also needed. Another tool is required in
order to devise for the extraction of results of the internal control mechanisms and their transformation and
storage in the previous component. This extraction of data provided by internal control mechanisms should be
made as soon as possible after the occurrence of the event monitored by the respective control.
A key requirement of this proposal is the development of risk profiles repository. It should be able to
maintain and manage the known negative and positive profiles of each organizational transaction to be
monitored and audited. These profiles must be also modeled according to the ontological model. Another key
requirement is the development of a module which compares the data from the internal control mechanisms
with the records maintained in the risk profiles repository. In addition, it must be able to determine which
profile is being followed by running each transaction. Moreover, if the situation that is running is unknown
and is not classified in the risk profiles repository, it is intended that this module is capable of introducing this
194 Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198

new behavior in the repository. Then, it should require the classification of this new profile by the responsible
of the system, now designated as transactional auditor.
Finally, the on-line results of the comparison module presented above should be stored in a repository in
real time. This repository will contain the history of the results of the transaction auditing and monitoring and
a picture of the current situation regarding the organizational transactions still in progress. This component
will allow an interface with the transactional auditor, queries, the preparation of audit reports, notifications
and alerts.
The architecture represented schematically in Fig. 1. was conceptualized based on the sought objectives
and in the general requirements specified. From the analysis of this architecture, it is perceived that the
proposed solution is intended to be permanently connected to the organizations operational information
system. In other words, the internal control mechanisms should be incorporated in the ERP in order to monitor
the status of the various phases and stages defined by the ontological model of organizational transactions.
Thereby, they are able to support the proposed solution, and consequently the monitoring and auditing of
organizational transactions.
The component represented by number 8 is the element responsible for the extraction of the state of
internal control mechanisms and the data they may provide in order to integrate all this information in a
repository (component 1). This will manage and maintain this information referring to the various states of
execution of transactions. Component 2 illustrates the risk profiles repository of the organizational
transactions to be monitored and audited.
The component represented by number 3 is the module aimed to be able to compare the various records in
the risk profiles repository (via the data flow 6) and determine which profile is being followed by running
each transaction, according to information received by component 1 (through the data flow 5). Furthermore, if
we are facing a situation which is not classified in the risk profiles repository, it is intended that this module is
capable of introducing this new behavior in the repository (via the data flow 7) and subsequently be classified
as positive or negative profile. The results carried out by the comparison module (component 3) are sent to a
repository (component 4), on which an interface for viewing current and historic state of controlling of the
audited transactions should be developed. The auditing repository should notify and alert the transactional
auditor when a negative behavior occurs, e.g. sending an e-mail or a sms.
Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198 195

Fig. 1. Conceptual architecture of the proposed solution

3.1. Technical considerations

CEP would be tendentiously the choice in order to develop the component responsible for the detection and
identification in real time of the risk profile. This, considering the set of facts presented in the literature review
and the fact that the risk profile is being followed in execution of the various organizational transactions to be
monitored and audited. Such a choice would be due to the features and performance of this paradigm in the
processing of large amount of events and its ability to respond in real time. However, there is a requirement of
the prototype which indicates that it must work with the organizations operational information systems in a
non-intrusive way, which puts into question, in some part, the use of CEP in the development of this
component. CEP is primarily designed to work with transient data, and data processing is done in memory. In
turn, the fact that the system must be non-intrusive, the functional architecture is designed so that it acts upon
data resulting from the execution of transactions in operational systems. In other words, the data to be
processed will be persistent.
Databases are an option to process persistent data, with the advantage that they do not have specified time
intervals contrary to the CEP processing. Because the system acts directly in the database of the operational
systems, it means that the organizational events associated with persistent data have already occurred. Thus,
the use of triggers in operational databases is a way of detection of events, since the insertion, edition or
deletion of a record or a record field means the occurrence of an event of a given organizational transaction.
Then, the component will render the activation of these triggers as if they were an occurrence of an event.
The "real time" is advertised as one of the requirements of the system and is defined within this work as the
time interval closer to the occurrence of an event, respecting the rhythm of execution of organizational
transactions. Thus, it seems that a database approach is sufficient to the functions of monitoring and auditing,
despite not having a lower latency as CEP. This, because the purpose of the system is not to act in a direct and
intrusive way on organizational execution of transactions, but rather to work with reports and alerts. Consider
196 Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198

that the time and the rhythm of organizational transactions are variable and in different orders of greatness (the
same transaction may have running times in the order of minutes, days or several months depending on the
situation in question).Therefore, the time in which the user of the prototype has to react in a corrective way,
after being alerted of an anomaly, is also variable, depending on the pace of the transaction in question. "A
real-time system is one, where the correctness not only depends on the functionality but also on the timeliness
of this functionality" [25].

4. Methodology and results

The deployment of the solution proposed above will intend to achieve some results. The first one is to
ascertain whether it is feasible to build a repository of risk profiles, following the structure and pattern of the
transaction axiom of Dietz's ontology, managing multiple positive and negative profiles of organizational
transactions. Another intended result is to demonstrate that the repository of risk profiles is a crucial element
in the real time monitoring and controlling of organizational transactions, checking if this repository has the
information needed for an analysis and evaluation of how transactions are being carried out.
The development and implementation of internal control mechanisms capable of providing information
about the transaction state in its various phases defined in Dietz's ontology are another result to be reached.
Finally we intended to attest that the proposed solution, implemented in accordance with the vision presented
in the research problem, is able to respond in real time about the state of execution of organizational
transactions, thus constituting a system with Continuous Assurance.
Based on the research problem presented, the solution projected and the results envisioned, triangulation is
the research approach proposed, followed by a qualitative approach, combined with aspects of a quantitative
approach [26, 27]. The choice of a qualitative approach is due to the fact that the research in question was
more targeted, in a general way, to aspects of management and organizations [28]. However, the quantitative
approach is justified because there will be a deployment of a prototype able to provide the collection and
analysis of some data that may lead to findings from the technological point of view [26].
For the classification of the interpretation of research, the positivist and interpretive epistemologies are the
ones to be used. The positivist epistemology will be used to objectively observe and analyze the results of
scientific investigation from the technological point of view [29]. Simultaneously, the interpretive
epistemology will be used to validate the resolution of the problem and understand the value of this result to
an organizational environment. However, the final interpretation is a partial analysis because it will be based
on a limited set of organizational transactions, subject of study [26, 30].
To conduct the research, the case study is the methodology which best suits the problem presented, because
the research is more empirical, investigating the feasibility of a system prototype in a real (simulated) context
and the resolution of organizational problems [26, 31]. Finally, the observation seems to be the appropriate
research technique/ tool to validate the raised research hypotheses. This technique is based on the observation
of a set of phenomena in order to collect data, on a systematic basis, about the behavior of the prototype. The
combination of indirect and direct observation seems to yield interesting results, by confrontation of users'
opinion with researchers' opinion [32, 33].
To concretize the case study we intend to deploy the prototype and evaluate it in an organizational
environment, and to this end, we aim to use the curriculum unit "Enterprise Simulation" from the Higher
Institute of Accounting and Administration of the University of Aveiro. This yields a controlled environment,
and also allows the application of the prototype in different organizational areas. "Enterprise Simulation",
included in the last year of the degree in Accounting, aims to simulate the organizational activities. These
activities incite dozens of groups of students to create their own enterprise (in one of various organizational
areas, like services, commerce, industry and public services), develop its operations in the business during an
Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198 197

operational period in accordance with the economic calendar and prepare and disseminate financial statements
[34]. The university provides a well-structured simulation environment that is very close to reality and an
infrastructure of information systems and that covers all the needs of organizations in their business activities.
The internal control mechanisms described in the architecture of the solution proposal will be incorporated in
this infrastructure of information systems which the university provides.

5. Final considerations

A solution with assurance services capable of continuous monitoring of organizational transactions in


compliance with the formalisms of a business ontological model is an innovative vision. This because
transactions are monitored and audited at a very low level, contrary to what happens in most monitoring of
transactions that occurs at a high level (for example, comparing whether a completed transaction followed a
set of established procedures). Another innovative vision presented in this paper is the implementation of a
repository that contains and maintains the risk profiles of the transactions to monitor and audit, following the
presented ontology.
This paper contributes for this new vision with a proposal of a conceptual architecture of a management
information system which aims continuous monitoring of organizational transactions executed and supported
exclusively in digital format supported by a business ontological model.

References
[1]. Markham, J. W., Financial history of modern United States corporate scandals Vol. 1. New York: M.E. Sharpe 2006
[2]. Verver, J. Risk Management and Continuous Monitoring. Retrieved March 1, 2011. Available from:
http://www.acl.com/pdfs/0303-Auditnet.pdf.
[3]. Bodoni, S. Kaupthing Creditors, Madoff, A-Tec, Lehman Brothers: Bankruptcy. Retrieved June 30, 2011. Available from:
http://www.bloomberg.com/news/2010-11-25/kaupthing-creditors-madoff-a-tec-lehman-brothers-bankruptcy.html.
[4]. Cohen, J., et al., Corporate Fraud and Managers Behavior: Evidence from the Press. In: Journal of Business Ethics, R. Cressy, D.
Cumming, C. Mallin, Editors. Springer Netherlands. 2011; p. 271-315.
[5]. Morais, M. G. C. T., A importncia da auditoria interna para a gesto: caso das empresas portuguesas, In 18 Congresso Brasileiro
de Contabilidade 2008: Gramado , Brazil. p. 1-15.
[6]. Vasarhelyi, M. A., M. Alles, K. T. Williams, Continuous assurance for the now economy. 1st ed. Sydney: Institute of Chartered
Accountants in Australia; 2010
[7]. PricewaterhouseCoopers, Internal Audit 2012. New York: PricewaterhouseCoopers LLP; 2007
[8]. Santos, C., Modelo Conceptual para Auditoria Organizacional Contnua com Anlise em Tempo Real. 1 ed: Editorial Novembro;
2009
[9]. Denning, D. E., An Intrusion-Detection Model. IEEE Transactions on Software Engineering, 1987. 13: p. 222-232.
[10]. Hepp, M.,D. Roman, An Ontology Framework for Semantic Business Process Management. Proceedings of Wirtschaftsinformatik,
2007.
[11]. Dietz, J. L. G., Enterprise Ontology: Theory and Methodology. New York: Springer-Verlag Inc.; 2006
[12]. Albani, A.,J. L. G. Dietz, Enterprise ontology based development of information systems. International Journal of Internet and
Enterprise Management, 2011. 7(1): p. 41-63.
[13]. McCoy, D. W., Business activity monitoring: Calm before the storm, In Technical Report LE-15-9727. 2002, Gartner.
[14]. Brandl, H.-M.,D. Guschakowski, Complex Event Processing in the context of Business Activity Monitoring - An evaluation of
different approaches and tools taking the example of the Next Generation easyCredit.Diploma Thesis. In Faculty Information
Technology/Mathematics. University of Applied Sciences Regensburg: Regensburg, 2007
[15]. ter Hofstede, A., W. van der Aalst, M. Weske, Business Process Management: A Survey. Business Process Management. M.
Weske, Editor. 2678; Springer Berlin / Heidelberg. 2003; p. 1019-1019.
[16]. Oliveira, J. J. R. d., Descoberta de Processos em Tempo Real.Master Dissertation. In Instituto Superior Tcnico. Univesidade
Tcnica de Lisboa: Lisboa, 2011
[17]. Roth, H., et al. Event data warehousing for Complex Event Processing. In Research Challenges in Information Science (RCIS),
2010 Fourth International Conference on. 2010.
[18]. Abadi, D. J., et al., Aurora: a new model and architecture for data stream management. The VLDB Journal The International
Journal on Very Large Data Bases 2003. 12(2): p. 120-139.
198 Rui Pedro Marques et al. / Procedia Technology 5 (2012) 190 198

[19]. Balazinska, M., H. Balakrishnan, M. Stonebraker, Load management and high availability in the Medusa distributed stream
processing system, In Proceedings of the 2004 ACM SIGMOD international conference on Management of data. 2004, ACM:
Paris, France. p. 929-930.
[20]. Abadi, D., et al. The Design of the Borealis Stream Processing Engine. In 2nd Biennial Conference on Innovative Data Systems
Research (CIDR'05). 2005.
[21]. Arasu, A., et al., STREAM: The Stanford Data Stream Management System. IEEE Data Engineering Bulletin Arasu, A, 2004.
26(1): p. 19 -26.
[22]. Greiner, T., et al., Business Activity Monitoring of norisbank Taking the Example of the Application easyCredit and the Future
Adoption of Complex Event Processing (CEP), In Proceedings of the IEEE Services Computing Workshops. 2006, IEEE Computer
Society. p. 83.
[23]. Ferreira, D. R.,D. Gillblad, Discovering Process Models from Unlabelled Event Logs, In Proceedings of the 7th International
Conference on Business Process Management. 2009, Springer-Verlag: Ulm, Germany. p. 143-158.
[24]. Ferreira, D. R., S. Alves, L. H. Thom, Ontology-Based Discovery of Workflow Activity Patterns, In 2nd International Workshop
on Reuse in Business Process Management. 2011: Clermont-Ferrand, France.
[25]. Mller, M. O., Structure and Hierarchy in Real-Time Systems.PhD Thesis. In Department of Computer Science. University of
Aarhus: Aarhus (Denmark), 2002
[26]. Grilo, R. M. M., Investigao em Sistemas de Informao Organizacionais em Portugal.Master Dissertation. In Departamento de
Engenharias. Universidade de Trs-Os-Montes e Alto Douro: Vila Real, 2008
[27]. Patton, M. Q., Qualitative Evaluation and Research Methods. 3 ed, ed. I. Sage Publications. Newbury Park, CA: Sage Publications,
Inc.; 2002
[28]. Myers, M. D. Qualitative Research in Information Systems. Retrieved March 1, 2011. Available from:
http://www.qual.auckland.ac.nz/.
[29]. Babbie, E., The practice of social research. Belmont, CA: Wadsworth Publishing Company; 1993
[30]. Giorgi, A., The theory, practice and evaluation of the phenomenological method as a qualitative research procedure. Journal of
Phenomenological Psychology, 1997. 28: p. 235-260.
[31]. Yin, R. K., Case Study Research, Design and Methods, ed. I. SAGE Publications. Vol. 5. Newbury Park: Sage Publications; 2009
[32]. Coolican, H., Research Methods and Statistics in Psychology: Hodder and Stoughton; 2004
[33]. Quivy, R.,L. V. Campenhoudt, Manual de Investigao em Cincias Sociais. Lisbon: Gradiva; 1998
[34]. Silva, F., As Tecnologias da Informao e Comunicao e o Ensino da Contabilidade.Master Dissertation. In Departamento de
Economia, Gesto e Engenharia Industrial. University of Aveiro, Portugal: Aveiro, 2009